Abstract: Water-repellent soil has a very wide application prospect in the field of agricultural engineering. However, it is still lacking in the hydraulic stability of water-repellent clay. The applications are only targeted at water-repellent dry sand so far. In this study, octadecylamine was taken to modify the highly adsorbent expansive soil with water repellency. A systematic investigation was also conducted to explore the water-repellent and hydraulic stability of loose expansive soil with different water repellency after adsorbing water vapor. Five samples were selected as the compacted expansive soil (hydrophilic expansive soil, hydrophilic expansive soil + hydrophilic sand, hydrophilic expansive soil + water repellent sand, water repellent expansive soil, water repellent expansive soil + water repellent sand). The results show that the greater the humidity was, the greater the mass of water vapor adsorbed by different water repellency expansive soil was. At the same humidity, the water repellency shared little effect on the mass adsorbed by soil. The expansive soil with different water repellency (including slight water repellency) was extremely water repellent after adsorption, where all test points were evenly water repellent. The huge specific surface area of water-repellent expansive soil was obtained for the strong adsorption properties. As such, the water repellency of expansive soil further increased to fully adsorb octadecylamine molecules while adsorbing water vapor during adsorption. After compaction, the water-repellent expansive soil and water-repellent expansive soil + water repellent sand samples exhibited water repellency in a large range of moisture content. There was a range of water-repellent moisture content, all of the soil samples were extremely water-repellent. Compared with the rest, the best hydraulic performance was achieved in the water-repellent expansive soil + water-repellent sand mixed sample share. Specifically, the expansion rate of the expansive soil was reduced from 33.9% to 3.45%, and the shrinkage fissure degree was reduced from 18.75% to 3.6%. Hydrophilic expansive soil realized 90% of the expansion within 90 min, while it took 1 680 min for water-repellent expansive soil and water-repellent sand to complete the same proportion of expansion. Water-repellent modification of expansive soil reduced the expansion amount and rate. There was the basically same air entry value of the hydrophilic sample and the water-repellent sample. But the water content corresponding to the air entry value of the water-repellent sample was smaller than that of the water-repellent sample, and the water-repellent expansive soil matric suction followed the same trend as hydrophilic expansive soil, in terms of moisture content. The pore water pressure remained "negative" at all times in the extremely water-repellent expansive soil, as the moisture content changed. There was no "positive" pore water pressure. There were the lower strength parameters c and φ of water-repellent expansive soil, compared with hydrophilic expansive soil. There was no significant change, indicating better mechanical properties after water-repellent treatment. The special hydraulic properties of water-repellent expansive soil were attributed to that the water-repellent aggregates remained dry inside the aggregates after immersion. This finding can provide a new idea for the application and promotion of water-repellent clay in agricultural engineering. An experimental basis can be offered to fully expand the application scope of water-repellent soil.